CN114186727B - Multi-cycle logistics network planning method and system - Google Patents

Abstract

The invention relates to a multi-period logistics network planning method and system. The method includes: acquiring business information in a logistics planning database; the business information includes demand flow direction, node information, line information, time-limitation requirement information and vehicle information; The business information is preprocessed to obtain model input data; the planning model is used to plan according to the model input data to obtain a logistics planning scheme; the logistics planning scheme includes an aging plan, a sorting plan, a route plan and a vehicle plan; according to the The logistics planning scheme determines the final logistics implementation planning scheme. The invention can improve the planning efficiency under the condition of obtaining the global optimal solution of network path planning.

Description

A multi-cycle logistics network planning method and system

technical field

The invention relates to the field of logistics path planning, in particular to a multi-period logistics network planning method and system.

Background technique

In order to solve the network planning problem under the complex network structure, the existing network planning methods either cannot obtain the global optimal solution, or consume a lot of computing resources. For example, based on the shortest path algorithm, the route of each flow direction is found, and then each optimization is only optimized for the sorting times and transportation distance of a single flow direction. Since the routing and line optimization of different flow directions are independent, the shared resources of different flow directions are not considered. In the actual implementation process of the optimized results, there is a problem that the route or vehicle plan between different transportation demand flows cannot meet the actual vehicle load or the minimum opening standard of the route. The solution based on each flow path, this scheme is based on the candidate basic network structure, adopts the enumeration method based on business constraints, and enumerates the feasible routes of all flow directions. candidate feasible routes. However, as the network scale increases, exhausting all alternative routes in all directions will consume a lot of computing resources, and even the problem of not being able to obtain all alternative routes in the actual scene, and the increase of alternative routes makes it difficult to solve the subsequent mathematical model. Exponential increase, the inability to quickly find a feasible solution in a limited time and other constraints. Even if the scale of routing exhaustion can be limited by some business rules, it will lead to the inability to obtain the global optimal solution in engineering applications and reduce the optimization space. Therefore, there is a need for a method that can find the global optimal solution for network planning without consuming a lot of computing resources.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a multi-period logistics network planning method and system, so as to improve the planning efficiency when the global optimal solution of network path planning is obtained.

For achieving the above object, the present invention provides the following scheme:

A multi-cycle logistics network planning method, comprising:

Obtain business information in the logistics planning database; the business information includes demand flow, node information, route information, time-limitation requirement information and vehicle information;

Preprocessing the business information to obtain model input data;

According to the input data of the model, the planning model is used for planning, and a logistics planning scheme is obtained; the logistics planning scheme includes an aging plan, a sorting plan, a route plan and a vehicle plan;

The final logistics implementation planning scheme is determined according to the logistics planning scheme.

Optionally, before the obtaining of the business information in the logistics planning database, the method further includes:

Acquiring business information in the operation database; the business information includes line information, node information and vehicle information, as well as demand flow and timeliness requirement information of the forecasting system;

Synchronizing the route information, the node information, the vehicle information, the demand flow direction and the time limit requirement information to the logistics planning database.

Optionally, the preprocessing of the business information to obtain model input data specifically includes:

Preprocessing the business information according to space-time nodes to obtain a set of space-time nodes;

Preprocessing the business information according to the time arc to obtain a time arc set;

Preprocessing the business information according to the transport arc to obtain a transport arc set;

Preprocessing the business information according to the demand flow to obtain a demand flow set;

Model input data is determined from the set of space-time nodes, the set of time arcs, the set of transport arcs, and the set of demand flows.

Optionally, the planning model includes an objective function and constraints;

The expression of the objective function is:

为需求流向i流向货量经过j段弧的百分比，cost_j为单位运输成本，sort_cost_j为单位分拣成本，m为需求流向i的上限，n为运输弧j的上限；Among them, od is the demand flow set, transport_arc is the transport arc set,

is the percentage of the demand flow direction i flow direction through the j segment arc, cost _j is the unit transportation cost, sort_cost _j is the unit sorting cost, m is the upper limit of the demand flow direction i, and n is the upper limit of the transportation arc j;

The expression of the constraint condition is:

为需求流向i流向货量经过j段弧的百分比，demand_i为需求流向i需要配送的货量，j为从节点d发出的运输弧，i为经过运输弧j的需求流向，node为时空节点集合，s为需求流向始发节点，t为需求流向目的节点，o为运输弧始发节点，d为运输弧目的节点，arc为运输弧和时间弧构成的网络弧集合，capacity_j为运输弧j的装载容量，x_j为运输弧j的频次，即为执行运输弧j所代表的运输计划使用车的辆数。in,

is the percentage of the demand flow through the j segment arc, demand _i is the quantity of goods that needs to be delivered to the demand flow i, j is the transportation arc sent from node d, i is the demand flow through the transportation arc j, node is the space-time node Set, s is the origin node of the demand flow, t is the destination node of the demand flow, o is the origin node of the transport arc, d is the destination node of the transport arc, arc is the network arc set composed of the transport arc and the time arc, and capacity _j is the transport arc The loading capacity of j, x _j is the frequency of the transport arc j, that is, the number of vehicles used to execute the transport plan represented by the transport arc j.

Optionally, the final logistics implementation planning scheme is determined according to the logistics planning scheme, which specifically includes:

Determine whether the logistics planning scheme satisfies the operation index; if yes, then determine that the logistics planning scheme is the final logistics implementation planning scheme; if not, use the logistics planning scheme to update the logistics planning database and the planning model , and return to step "Get business information from logistics planning database".

A multi-cycle logistics network planning system, comprising:

The first acquisition module is used to acquire business information in the logistics planning database; the business information includes demand flow direction, node information, route information, time-limitation requirement information and vehicle information;

a preprocessing module for preprocessing the business information to obtain model input data;

a planning module, used for planning by using the planning model according to the model input data to obtain a logistics planning scheme; the logistics planning scheme includes an aging plan, a sorting plan, a route plan and a vehicle plan;

The final logistics implementation planning scheme determination module is used for determining the final logistics implementation planning scheme according to the logistics planning scheme.

Optionally, also include:

The second acquisition module is used to acquire business information in the operation database; the business information includes line information, node information and vehicle information, as well as demand flow and timeliness requirement information of the forecasting system;

The synchronization module is used for synchronizing the route information, the node information, the vehicle information, the demand flow direction and the timeliness requirement information to the logistics planning database.

Optionally, the preprocessing module specifically includes:

a spatiotemporal node set determination unit, configured to preprocess the business information according to the spatiotemporal nodes to obtain a spatiotemporal node set;

a time arc set determining unit, configured to preprocess the business information according to the time arc to obtain a time arc set;

a transport arc set determining unit, configured to preprocess the business information according to the transport arc to obtain a transport arc set;

a demand flow direction set determination unit, configured to preprocess the business information according to the demand flow direction to obtain a demand flow direction set;

A model input data determining unit, configured to determine model input data according to the space-time node set, the time arc set, the transportation arc set, and the demand flow set.

Optionally, the planning model includes an objective function and constraints;

The expression of the objective function is:

为需求流向i流向货量经过j段弧的百分比，cost_j为单位运输成本，sort_cost_j为单位分拣成本，m为需求流向i的上限，n为运输弧j的上限；Among them, od is the demand flow set, transport_arc is the transport arc set,

is the percentage of the demand flow direction i flow direction through the j segment arc, cost _j is the unit transportation cost, sort_cost _j is the unit sorting cost, m is the upper limit of the demand flow direction i, and n is the upper limit of the transportation arc j;

The expression of the constraint condition is:

为需求流向i流向货量经过j段弧的百分比，demand_i为需求流向i需要配送的货量，j为从节点d发出的运输弧，i为经过运输弧j的需求流向，node为时空节点集合，s为需求流向始发节点，t为需求流向目的节点，o为运输弧始发节点，d为运输弧目的节点，arc为运输弧和时间弧构成的网络弧集合，capacity_j为运输弧j的装载容量，x_j为运输弧j的频次，即为执行运输弧j所代表的运输计划使用车的辆数。in,

is the percentage of the demand flow through the j segment arc, demand _i is the quantity of goods that needs to be delivered to the demand flow i, j is the transportation arc sent from node d, i is the demand flow through the transportation arc j, node is the space-time node Set, s is the origin node of the demand flow, t is the destination node of the demand flow, o is the origin node of the transport arc, d is the destination node of the transport arc, arc is the network arc set composed of the transport arc and the time arc, and capacity _j is the transport arc The loading capacity of j, x _j is the frequency of the transport arc j, that is, the number of vehicles used to execute the transport plan represented by the transport arc j.

Optionally, the final logistics implementation planning scheme is determined according to the logistics planning scheme, which specifically includes:

Determine whether the logistics planning scheme satisfies the operation index; if yes, then determine that the logistics planning scheme is the final logistics implementation planning scheme; if not, use the logistics planning scheme to update the logistics planning database and the planning model , and return to step "Get business information from logistics planning database".

According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

The present invention obtains the business information in the logistics planning database; the business information includes demand flow direction, node information, route information, aging requirement information and vehicle information; the business information is preprocessed to obtain model input data; The input data is planned using the planning model to obtain a logistics planning scheme; the final logistics implementation planning scheme is determined according to the logistics planning scheme. Compared with the path-based global network planning, the exhaustive routing process is omitted, the data processing process is simplified, and the calculation efficiency of the planning model is improved. And there is no need to delete routes based on business rules, so that the final logistics implementation planning scheme is the global optimal solution under the planning constraints.

Description of drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings required in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some of the present invention. In the embodiments, for those of ordinary skill in the art, other drawings can also be obtained according to these drawings without creative labor.

Fig. 1 is the flow chart of the multi-cycle logistics network planning method provided by the present invention;

Figure 2 is a schematic diagram of a network planning scheme;

3 is a schematic diagram of a multi-cycle logistics network planning method provided by the present invention;

FIG. 4 is a schematic structural diagram of a multi-cycle logistics network planning system provided by the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

The purpose of the present invention is to provide a multi-period logistics network planning method and system, so as to improve the planning efficiency when the global optimal solution of network path planning is obtained.

In order to make the above objects, features and advantages of the present invention more clearly understood, the present invention will be described in further detail below with reference to the accompanying drawings and specific embodiments.

As an important means to ensure the efficient operation of the logistics network, network planning aims to complete the transportation tasks in a specific flow direction by rationally arranging routes, capacity and routing plans. There are two basic concepts in the network planning problem, namely the basic network structure and the flow of transportation demand. Among them, Figure 2(a) is a network planning problem, and Figure 2(b) is a feasible network planning scheme. As shown on the left side of the dotted line in Figure 2(a), Beijing, Xi'an, Wuhan and Shenzhen can form a simple network. Candidate lines can be opened between any two points, and all open candidate lines and nodes constitute the basic network structure. As shown on the right side of the dotted line in Figure 2(a), <Beijing-Shenzhen, 50 tons>, <Beijing-Wuhan, 20 tons>, <Wuhan-Shenzhen, 30 tons> and <Xi’an-Shenzhen, 10 tons> are 4 A transport demand flow. How to select the appropriate line combination with the goal of minimizing the cost, and complete the distribution tasks of the four demand flows under the premise of satisfying the actual constraints, belongs to the classic scenario in network planning. As shown in Figure 2(b), many possible line combinations can be found. In Figure 2(b), both solution 1 and solution 2 are feasible network structures, but the line combination of solution 2 is shorter because of fewer lines, and the number of feasible solutions in the actual network is less than It is a challenging task to select the optimal network structure from a large number of feasible solutions.

As shown in Figure 1, a multi-period logistics network planning method provided by the present invention includes:

Step 101: Acquire business information in the logistics planning database; the business information includes demand flow direction, node information, route information, time-limitation requirement information and vehicle information.

Step 102: Preprocess the business information to obtain model input data. Step 102 specifically includes: preprocessing the business information according to space-time nodes to obtain a set of space-time nodes; preprocessing the business information according to time arcs to obtain a time arc set; preprocessing the business information according to transport arcs. processing to obtain a transport arc set; preprocessing the business information according to the demand flow direction to obtain a demand flow set; determining a model according to the space-time node set, the time arc set, the transport arc set and the demand flow set Input data. The time arc represents the arc connecting two different time nodes on the same space node, the arc length represents a period of time on the space node, and the capacity of the arc represents the sorting capacity in this time period. The transport arc represents the arc connecting two time nodes on different space nodes, the length of the arc is transferred from one space node to another space node, and the capacity of the arc represents the vehicle (such as a vehicle or a vehicle) represented by the transport arc. the loading capacity of an aircraft, etc.).

Step 103 : perform planning by using a planning model according to the model input data to obtain a logistics planning scheme; the logistics planning scheme includes an aging plan, a sorting plan, a route plan and a vehicle plan. The planning model in the present invention is an edge-flow based network flow model (arc-flow based network flow model, arc-flow model).

Wherein, the planning model includes an objective function and constraints;

其中，od为需求流向集合，transport_arc为运输弧集合，

为需求流向i流向货量经过j段弧的百分比，cost_j为单位运输成本，sort_cost_j为单位分拣成本，m为需求流向i的上限，n为运输弧j的上限；The expression of the objective function is:

Among them, od is the demand flow set, transport_arc is the transport arc set,

is the percentage of the demand flow direction i flow direction through the j segment arc, cost _j is the unit transportation cost, sort_cost _j is the unit sorting cost, m is the upper limit of the demand flow direction i, and n is the upper limit of the transportation arc j;

The expression of the constraint condition is:

为需求流向i流向货量经过j段弧的百分比，demand_i为需求流向i需要配送的货量，j为从节点d发出的运输弧，i为经过运输弧j的需求流向，node为时空节点集合，s为需求流向始发节点，t为需求流向目的节点，o为运输弧始发节点，d为运输弧目的节点，arc为运输弧和时间弧构成的网络弧集合，capacity_j为运输弧j的装载容量，x_j为运输弧j的频次，即为执行运输弧j所代表的运输计划使用车的辆数。in,

is the percentage of the demand flow through the j segment arc, demand _i is the quantity of goods that needs to be delivered to the demand flow i, j is the transportation arc sent from node d, i is the demand flow through the transportation arc j, node is the space-time node Set, s is the origin node of the demand flow, t is the destination node of the demand flow, o is the origin node of the transport arc, d is the destination node of the transport arc, arc is the network arc set composed of the transport arc and the time arc, and capacity _j is the transport arc The loading capacity of j, x _j is the frequency of the transport arc j, that is, the number of vehicles used to execute the transport plan represented by the transport arc j.

Step 104: Determine a final logistics implementation planning scheme according to the logistics planning scheme.

Step 104 specifically includes:

Determine whether the logistics planning scheme satisfies the operation index; if yes, then determine that the logistics planning scheme is the final logistics implementation planning scheme; if not, use the logistics planning scheme to update the logistics planning database and the planning model , and return to step 101. Using the logistics planning scheme to update the planning model is to use the logistics planning scheme to increase the constraints in the planning model, so as to make the planning model more realistic. Among them, the operation indicators refer to the key operation indicators output through the simulation model, such as whether the sorting volume of the transfer yard under the new scheme exceeds the design peak, and whether the loading and unloading tasks of each platform cause too many unloading slots due to the new scheme. The queuing scene and so on are closer to realistic and detailed indicators.

In a practical application, before step 101, the method further includes: acquiring business information in the operation database; the business information includes line information, node information and vehicle information, as well as the demand flow and timeliness requirement information of the prediction system; , the node information, the vehicle information, the demand flow and the timeliness requirement information are synchronized to the logistics planning database.

As shown in FIG. 3 , the method provided by the present invention is specifically divided into six modules: planning database, data preprocessing, arc-flow model, scheme verification, operation database and planning requirement update. Among them, the operation system is an information management system that supports the normal operation of the enterprise logistics network, such as waybill management, routing management, sorting field management system, etc. This type of information management system manages routes, cargo distribution requirements, site information, etc. The data related to the network and customer needs are regularly deposited to form an operational database, which records the operation of the solution in practice. The planning database is used to provide basic business data for the data preprocessing of the arc-flow model, and its data records the information related to the current state of the network and the predicted distribution demand in the operating system. It mainly provides input data for the data preprocessing module. The data preprocessing module processes the business processing data in the planning database into a standard data set that can be recognized by the arc-flow model. The Arc-flow model is responsible for outputting a network solution that meets all distribution needs. The scheme verification module simulates and evaluates the cost, site, and route of the scheme under different volume levels in a way that is more in line with the actual state of the business, and verifies the robustness of the scheme in different scenarios. And according to the evaluation results, after updating the constraints of planning requirements, iterative optimization is continuously carried out according to the process of Figure 3 until certain conditions are met.

Step 1: The drawing module of the planning database synchronizes the basic information of various alternative routes, nodes and vehicle networks from the operation database, and obtains business demand information such as demand flow and customer timeliness requirements from the waybill or forecast system. This dataset is deposited into the planning database as a basic data source to solve network planning problems.

Step 2: The data preprocessing module converts the business information of the planning database into the standard model input data format required by the arc-flow model.

Step 3: Based on the network structure required for planning and business distribution requirements, the arc-flow model outputs the optimized route combination, flow routing and route plan solutions.

Step 4: In order to improve the output efficiency of the solution, the arc-flow model needs to abstract the actual problem and eliminate some secondary operating rules, such as the departure rules, the flow of goods in the sorting system and other more detailed information. Therefore, before the solution is synchronized to the operation system for implementation, the solution verification module needs to evaluate the feasibility and rationality of the solution based on simulation or manual experience. For example, a network simulation model containing more operational details (such as departure rules, sorting processes in the transfer yard, etc.) can be established based on discrete event simulation technology. The simulation model receives the network solutions (generally including routes, vehicles and routing plans, etc.) from the optimized output of the arc-flow model, and evaluates the solutions output by the arc-flow model in a more realistic digital twin world. Observe whether there will be other operational problems, such as whether there are too long vehicle loading and unloading queues in some transit yards, and oversaturation caused by insufficient sorting capacity. If the solution is feasible, the sorting and line output of the arc-flow model can be Synchronized with the vehicle plan to the operation database, the operation system will generate the distribution route of each demand flow and the sorting plan of the corresponding node based on the optimized route, and prepare the vehicle, plane and railway in advance according to the planned vehicle plan based on the arc-flow model. capacity resources. If the plan is unreasonable, based on the evaluation results, some important constraints are updated, such as whether the oversaturated sorting can be transformed to expand the original sorting capacity. As a result, new planning requirements are generated. The planning requirements data generated by the assessment pass through the planning requirements update module to update the information in the planning database, and then go through the data preprocessing module to generate new model input data, and iteratively output the final implementation plan. .

Among them, some important constraints are updated based on the evaluation results, specifically, the constraints are updated first, and then updated to the arc-flow model for the next iteration, and a new optimization scheme is output. For example, after evaluating the network solution output in the previous step, it is found that some transfer yard platforms and sorting capabilities are insufficient. In the next step of optimization, constraints will be added to limit the departure and arrival of the transfer yard. Then output the next version of the new network solution.

The updating of the information in the planning database is specifically, for example, in terms of line information in the planning database. Before optimization, according to the network structure synchronized from the operational database, 1,000 lines are needed to ensure that all transportation needs can be delivered. After the optimization of our method, we need to delete 50 lines from the original line and add 10 lines. Only 960 lines are needed to meet the distribution needs, and the cost is cheaper. We will first synchronize the optimized route information to the planning database, and then perform manual or simulation verification and evaluation, through iteration until the plan is executable, and then synchronize from the planning database to the operation database to form a new route, sorting and vehicle arrangement plan .

Iteratively output the final feasible solution specifically for the Arc-flow model output solution, the solution checks the model to verify the feasibility of the solution, generates new business constraints, and then adds the new constraints to the arc-flow model, outputs a new network solution, and re- To verify the feasibility of the program. For example, iterate continuously until the final output can be implemented.

In FIG. 3 , since the data preprocessing module and the arc-flow model module are two key modules in the process, and are also the content to be protected by this patent, the details of the two modules will be described in detail below. In the part of data preprocessing module, it includes four parts: time node, time arc, transportation arc and demand flow processing.

Tables 1-4 are the details of the model standard input data presented after the business data in the planning database is processed by the data preprocessing module. This data will be used as a model input that the arc-flow model can recognize. Table 1 shows the time nodes in the space-time network diagram, Table 2 shows the time arcs in the space-time network diagram, Table 3 shows the transportation arcs in the space-time network diagram, and Table 4 shows the demand flow in the space-time network diagram. It is well known that a network is generally composed of nodes and arcs. Tables 1-3 constitute the spatiotemporal network diagrams required by the arc-flow model. Table 1 shows the network nodes in the spatiotemporal network diagram. Table 2 and Table 3 represent the network arcs in the figure, and their function is to connect the network nodes in Table 1, so that the nodes in Table 1 are connected into a fully connected space-time network. Table 2 represents the connection arcs between any two adjacent time nodes on the same space node in Table 1, and Table 3 represents the network arcs connecting different network nodes on different space nodes in Table 1. The spatio-temporal network diagram is the mapping of transfer shifts, route data and transportation and distribution demand data in the actual network. From this, we can reduce the dimension of the real two-dimensional space-time network into a one-dimensional network. The example data of each table will be explained in detail below.

Table 1 Time nodes in the spatiotemporal network diagram

key value time frame date time space node TN001 1-0700-Beijing 1 07:00 Beijing TN002 1-0830-Beijing 1 08:30 Beijing TN003 1-1100-Beijing 1 11:00 Beijing TN004 1-0700-Shenzhen 1 07:00 Shenzhen TN005 2-1500-Wuhan 2 15:00 Wuhan TN006 3-2000-Shenzhen 3 20:00 Shenzhen

Table 1 is the standard data format of the time node, the key value represents the unique identifier, the time node represents the time information of each space node, and the space node in the table is Beijing 3 rows of data as an example. TN001, TN002, and TN003 represent three time nodes of 07:00, 08:30, and 11:00 on the Beijing node, respectively. This node represents the key time nodes in the operation such as line departure, line arrival, shift start and shift end.

Table 2 Time arcs in the spatiotemporal network diagram

Table 2 represents the virtual time arcs connecting two adjacent time nodes on the same space node, TA001 and TA002 represent 2 virtual time arcs connecting three time nodes on the Beijing node at 07:00, 08:00 and 11:00 . For example, TA001 in Table 2 represents the arc of two time nodes TN001 and TN002 in the table. It should be noted that the time arc connects two different time nodes of the same space node. For example, TN003 and TN004 in Table 1 cannot form a time arc because they are time nodes on two different spatial nodes in Beijing and Shenzhen. It should be noted that the processing capacity of this time arc represents the sorting capacity or the unloading capacity within a specified time.

Table 3. Transport arcs in the spatiotemporal network diagram

Table 3 represents the time arcs connecting time nodes on different spatial nodes, which we call transport arcs. It represents the lines present in the candidate network structure. As shown in Table 3, TP001 and TP002 represent the two lines connecting Beijing-Wuhan and Wuhan-Shenzhen respectively. Taking TP001 as an example, this transport arc connects the two network nodes TN001 and TN005 in Table 2. It means that there is a line from Beijing to Wuhan in the actual logistics network. The departure time of this line in Beijing is 07:00 on the first day and the arrival time in Wuhan is 15:00 on the second day. The loading capacity represents the vehicle loading capacity of the designated line, indicating that there is 150 tons of available load space on this line. The cost of the transport arc is the line unit cost.

Table 4 Demand flow in the spatiotemporal network diagram

Table 4 shows the standard data structure of the transportation demand flow in the network. DO001 and DO002 represent the two distribution demands of Beijing-Shenzhen and Wuhan-Shenzhen. Among them, DO001 indicates that the TN001 node has a supply of 10 tons in the space-time network diagram, and the TN006 node has a demand of 20 tons. It means that there is a batch of goods from Beijing destined for Shenzhen at 07:00 on the first day in the actual network, and it needs to arrive in Shenzhen before 20:00 on the third day.

After the above data preprocessing, all the model standard input data required to build the arc-flow model are obtained. The objective function of the arc-flow model is,

This objective function represents the minimization of network transportation costs.

Constraints are

In the first constraint, d≠s&d≠t indicates that for any non-demand originating and destination nodes, the traffic is balanced.

In the second constraint, d=s, it means that for any demand-originating node, the supply must be fully delivered.

In the third constraint, d=t, it means that for any demand destination node, all demands must be satisfied.

The fourth constraint states that for any line, the line full capacity constraint must be satisfied.

Among them, node is a node collection composed of space-time nodes, and the value of the node collection is the collection shown in the key value field in Table 1. Contains TN001, TN002, TN003, etc.

time_arc _{<i, j>} is an arc set composed of time arcs, where i, j∈node, the value of time_arc _{<i, j>} set is the set shown in the key value field in Table 2, as shown in the data example of Table 2 , TA001 and TA002 in Table 2 constitute the set of time_arc _{<i, j>} . where i, j represent the originating node and destination node of time_arc _{<i, j>} respectively.

transport_arc _{<i, j>} is the arc set consisting of transport arcs, where i, j∈node. The value of the transport_arc _{<i, j>} set is the set shown in the key value field in Table 3. As shown in the data example in Table 3, TP001 and TP002 in Table 2 constitute the set of transport_arc. where i, j represent the originating node and destination node of transport_arc _{<i, j>} respectively.

arc _{<i, j>} is a network arc set composed of transport arcs and time arcs, where arc _{<i, j>} = transport_arc _{<i, j>} ∪ time_arc _{<i, j>} transport arc.

od _{<i, j>} is the demand flow set composed of space-time nodes, and the value of the od _{<i, j>} set is the set shown in the key value field in Table 4, including DO001 and DO002. where i is the originating node and j is the destination node. And i, j∈node.

cost _i is the unit transportation cost, i ∈ transport_arc. The value of this parameter is shown in the cost field in Table 3. For example, cost _TP001 = 1920, indicating that the cost of the transport arc TP001 is 1920 yuan per trip.

capacity _i : If i∈transport_arc, it is the full capacity of the line. The value of this parameter is shown in the loading capacity field in Table 3. For example, capacity _TP001 = 150, indicating that the maximum loading capacity of TP001 is 150 square meters. If i∈time_arc, it is the sorting capacity of this time period, the value of this parameter is shown in the processing capacity field in Table 3, such as capacity _TA001 = 1500, indicating that TA001 represents 07:00-08:30 to this time The maximum sorting capacity is 1500 cubic meters.

demand _i is the quantity of goods that needs to be delivered to demand direction i, i∈od, the value of this parameter is shown in Table 4. For example, demand _DO001 = 20, it means that the quantity of goods that needs to be delivered to DO001 is 20 square meters.

为决策变量，其中i∈od，j∈transport_arc，该决策变量表示i流向货量经过j段弧的百分比。

is the decision variable, where i∈od, j∈transport_arc, the decision variable represents the percentage of the flow of goods in the i direction through the j segment arc.

The arc-flow model uses solvers such as scip and cpelx to mathematically model and solve the above expressions.

The present invention first designs a new arc-flow model for network planning, which converts a multi-period two-dimensional space-time network into a one-dimensional space-time network, reduces the waste of computing resources caused by exhaustive routing in practical applications, and is the most efficient. The optimal solution is the global optimal solution. Secondly, a new method for solving practical network planning problems is proposed, combining optimization, simulation and real business systems, which can provide the function of a planning hub with real-time optimization for complex networks.

As shown in Figure 4, a multi-period logistics network planning system provided by the present invention is characterized in that, it includes:

The first obtaining module 201 is configured to obtain business information in the logistics planning database; the business information includes demand flow direction, node information, route information, time-limitation requirement information and vehicle information.

The preprocessing module 202 is configured to preprocess the business information to obtain model input data.

The planning module 203 is used for planning by using the planning model according to the model input data to obtain a logistics planning scheme; the logistics planning scheme includes an aging plan, a sorting plan, a route plan and a vehicle plan.

The final logistics implementation planning scheme determination module 204 is configured to determine the final logistics implementation planning scheme according to the logistics planning scheme.

In practical applications, it also includes:

The second acquisition module is used for acquiring business information in the operation database; the business information includes line information, node information and vehicle information, as well as demand flow and time-limitation requirement information of the forecasting system.

The synchronization module is used for synchronizing the route information, the node information, the vehicle information, the demand flow direction and the timeliness requirement information to the logistics planning database.

In practical applications, the preprocessing module 202 specifically includes: a spatiotemporal node set determination unit, configured to preprocess the service information according to spatiotemporal nodes to obtain a spatiotemporal node set; a time arc set determination unit, used for The business information is preprocessed according to the time arc to obtain the time arc set; the transport arc set determining unit is used for preprocessing the business information according to the transport arc to obtain the transport arc set;

A demand flow direction set determination unit, used for preprocessing the business information according to the demand flow direction, to obtain a demand flow direction set; a model input data determination unit, used for according to the space-time node set, the time arc set, the transport arc set The collection and the requirements flow to the collection determination model input data.

In practical applications, the planning model includes an objective function and constraints;

The expression of the objective function is:

为需求流向i流向货量经过j段弧的百分比，cost_j为单位运输成本，sort_cost_j为单位分拣成本，m为需求流向i的上限，n为运输弧j的上限；Among them, od is the set of demand flow directions, transport_arc is the set of transport arcs, and demand _i is the quantity of goods that needs to be delivered to the demand flow direction i,

is the percentage of the demand flow direction i flow direction through the j segment arc, cost _j is the unit transportation cost, sort_cost _j is the unit sorting cost, m is the upper limit of the demand flow direction i, and n is the upper limit of the transportation arc j;

The expression of the constraint condition is:

为需求流向i流向货量经过j段弧的百分比，j为从节点d发出的运输弧，i为经过运输弧j的需求流向，node为时空节点集合，s为需求流向始发节点，t为需求流向目的节点，o为运输弧始发节点，d为运输弧目的节点，arc为运输弧和时间弧构成的网络弧集合，capacity_j为运输弧j的装载容量，x_j为运输弧j的频次，为执行运输弧j所代表的运输计划使用车的辆数。in,

is the percentage of the demand flow through the j segment arc, j is the transport arc from node d, i is the demand flow through transport arc j, node is the collection of space-time nodes, s is the demand flow originating node, and t is The demand flows to the destination node, o is the originating node of the transport arc, d is the destination node of the transport arc, arc is the network arc set composed of the transport arc and the time arc, capacity _j is the loading capacity of the transport arc j, and x _j is the transport arc j's load capacity. Frequency, the number of vehicles used to execute the transportation plan represented by the transportation arc j.

In practical applications, the final logistics implementation planning scheme determination module 204 specifically includes:

A judging unit for judging whether the logistics planning scheme satisfies the operation index; if so, determining that the logistics planning scheme is the final logistics implementation planning scheme; if not, using the logistics planning scheme to analyze the logistics planning database and all Update the planning model described above, and return to step "Get business information in logistics planning database".

The present invention abstracts the two-dimensional space-time network into a one-dimensional space-time network including periods, space nodes and time nodes. Abstract attributes such as duration, processing capacity, parking space, sorting cost, and transportation cost into various attributes of arcs, and based on solvers such as scip, cplex or gurobi, obtain the routes, paths, sorting, and vehicle plans of the complex network, etc. Network planning solutions. Then use the method of simulation verification and evaluation to verify and evaluate the network solution obtained based on the arc-flow model, obtain the unreasonable route, sorting and vehicle plans in the solution, and then add such unreasonable solutions to arc-flow model constraints, and output an iterative network solution based on the updated arc-flow model until the solution meets practical application requirements. Compared with the path-based global network planning, this scheme omits the process of exhaustive routing, simplifies the data processing process, and improves the computational efficiency of the models of other schemes. And there is no need to delete routes based on business rules, so that the solution obtained is the global optimal solution under the planning constraints.

The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same and similar parts between the various embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant part can be referred to the description of the method.

In this paper, specific examples are used to illustrate the principles and implementations of the present invention. The descriptions of the above embodiments are only used to help understand the methods and core ideas of the present invention; meanwhile, for those skilled in the art, according to the present invention There will be changes in the specific implementation and application scope. In conclusion, the contents of this specification should not be construed as limiting the present invention.

Claims (8)

1. A multi-cycle logistics network planning method is characterized by comprising the following steps:

acquiring service information in a logistics planning database; the service information comprises a demand flow direction, node information, line information, timeliness requirement information and vehicle information;

preprocessing the service information to obtain model input data;

planning by using a planning model according to the model input data to obtain a logistics planning scheme; the logistics planning scheme comprises an aging plan, a sorting plan, a line plan and a vehicle plan;

determining a final logistics implementation planning scheme according to the logistics planning scheme;

the planning model comprises an objective function and a constraint condition;

the expression of the objective function is:

wherein od is the demand flow set, transport _ arc is the transport arc set,

percentage of the flow of demand to i to the flow of cargo through the j arc, cost _j For unit transportation cost, sort _ cost _j Taking the unit sorting cost as a unit, wherein m is the upper limit of the demand flow direction i, and n is the upper limit of the transport arc j;

the expression of the constraint condition is as follows:

wherein,

demand flow to i flow to the percentage of the volume of the cargo passing through the j arc, demand _i The quantity of goods to be distributed is required to flow to i in a demand flow direction, j is a transportation arc sent from a node d, i is the demand flow direction passing through the transportation arc j, node is a space-time node set, s is a demand flow direction starting node, t is a demand flow direction target node, o is a transportation arc starting node, d is a transportation arc target node, arc is a network arc set formed by the transportation arc and a time arc, capacity _j For the loading capacity of the transport arc j, x _j The frequency of the transport arc j, i.e., the number of vehicles used for the transport plan represented by the execution of the transport arc j.

2. The multi-cycle logistics network planning method of claim 1, further comprising, before the obtaining the service information in the logistics planning database:

acquiring service information in an operation database; the service information comprises line information, node information, vehicle information and demand flow direction and timeliness requirement information of a prediction system;

and synchronizing the line information, the node information, the vehicle information, the demand flow direction and the timeliness requirement information to the logistics planning database.

3. The multi-cycle logistics network planning method of claim 1, wherein the preprocessing the service information to obtain model input data specifically comprises:

preprocessing the service information according to space-time nodes to obtain a space-time node set;

preprocessing the service information according to a time arc to obtain a time arc set;

preprocessing the service information according to a transport arc to obtain a transport arc set;

preprocessing the service information according to a demand flow direction to obtain a demand flow direction set;

and determining model input data according to the space-time node set, the time arc set, the transportation arc set and the demand flow direction set.

4. The multi-cycle logistics network planning method of claim 1, wherein the determining a final logistics implementation planning scheme according to the logistics planning scheme specifically comprises:

judging whether the logistics planning scheme meets the operation index; if so, determining the logistics planning scheme as a final logistics implementation planning scheme; if not, updating the logistics planning database and the planning model by using the logistics planning scheme, and returning to the step of acquiring the service information in the logistics planning database.

5. A multi-cycle logistics network planning system, comprising:

the first acquisition module is used for acquiring service information in the logistics planning database; the service information comprises a demand flow direction, node information, line information, timeliness requirement information and vehicle information;

the preprocessing module is used for preprocessing the service information to obtain model input data;

the planning module is used for planning by using a planning model according to the model input data to obtain a logistics planning scheme; the logistics planning scheme comprises an aging plan, a sorting plan, a line plan and a vehicle plan;

the final logistics implementation planning scheme determining module is used for determining a final logistics implementation planning scheme according to the logistics planning scheme;

the planning model comprises an objective function and a constraint condition;

the expression of the objective function is:

wherein od is the demand flow set, transport _ arc is the transport arc set,

percentage of the flow of demand to i to the flow of cargo through the j arc, cost _j For unit transportation cost, sort _ cost _j Taking the unit sorting cost as a unit, wherein m is the upper limit of the demand flow direction i, and n is the upper limit of the transport arc j;

the expression of the constraint condition is as follows:

wherein,

demand flow to i flow to the percentage of the volume of the cargo passing through the j arc, demand _i The amount of goods to be delivered for a demand flow direction i, j is a traffic arc issued from a node d, i is the demand flow direction passing through the traffic arc j, node is a space-time node set, s is a demand flow direction originating node, t is a demand flow direction destination node, o is a traffic arc originating node, and d is a traffic arc destination nodePoint, arc, is a set of network arcs, capacity, formed by a transport arc and a time arc _j For the loading capacity of the transport arc j, x _j The frequency of the transport arc j, i.e., the number of vehicles used for the transport plan represented by the execution of the transport arc j.

6. The multi-cycle logistics network planning system of claim 5, further comprising:

the second acquisition module is used for acquiring the service information in the operation database; the service information comprises line information, node information, vehicle information and demand flow direction and timeliness requirement information of a prediction system;

and the synchronization module is used for synchronizing the line information, the node information, the vehicle information, the demand flow direction and the timeliness requirement information to the logistics planning database.

7. The multi-cycle logistics network planning system of claim 5, wherein the preprocessing module specifically comprises:

the space-time node set determining unit is used for preprocessing the service information according to space-time nodes to obtain a space-time node set;

the time arc set determining unit is used for preprocessing the service information according to a time arc to obtain a time arc set;

the transportation arc set determining unit is used for preprocessing the service information according to the transportation arcs to obtain a transportation arc set;

a demand flow direction set determining unit, configured to preprocess the service information according to a demand flow direction to obtain a demand flow direction set;

and the model input data determining unit is used for determining model input data according to the space-time node set, the time arc set, the transportation arc set and the demand flow direction set.

8. The multi-cycle logistics network planning system of claim 5, wherein the final logistics implementation planning scheme determination module specifically comprises:

the judging unit is used for judging whether the logistics planning scheme meets the operation index; if so, determining the logistics planning scheme as a final logistics implementation planning scheme; if not, updating the logistics planning database and the planning model by using the logistics planning scheme, and returning to the step of acquiring the service information in the logistics planning database.

Images